Quantum-Enhanced Brain Tumor Detection and Progression Prediction Using MRI Imaging

Brain tumor identification and change over time analysis are essential for timely diagnosis and effective treatment scheduling and planing. This study presents a hybrid quantum-classical deep learning framework integrating Quantum Convolutional Neural Networks (QCNNs) with classical CNN to improve MRI-based tumor classification. Unlike traditional CNNs, which suffer from high computational costs and limited feature extraction capabilities, the proposed Quantum-Enhanced Tumor Analysis Framework (QETAF) leverages quantum feature maps to enhance tumor localization and segmentation. This study utilizes the BraTS MRI dataset (comprising 67,000 labeled scans) and applies contrast enhancement, intensity normalization, and augmentation techniques for preprocessing. The novel hybrid model employs CNN model for extracting the essential features initially and QCNN for refined feature representation, significantly improving tumor classification accuracy. Moreover, morphological variations can be monitored using Recurrent Quantum Neural Networks (RQNNs), which have been employed to track tumor progression. According to investigational results, RQNN increases the accuracy of tumor progress prediction, whereas QCNN beats regular CNNs with an 89% Dice Coefficient. Compared to classical models, the proposed approach reduces inference time by 28% while maintaining superior classification performance. This quantum-assisted model presents a novel pathway for enhancing computational efficiency and precision in brain tumor diagnostics, covering the way for more consistent clinical diagnostics.